[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
20030416: McIDAS XCD/ADDE setup at UQAM (cont.)
- Subject: 20030416: McIDAS XCD/ADDE setup at UQAM (cont.)
- Date: Wed, 16 Apr 2003 16:56:58 -0600
>From: Unidata Support <address@hidden>
>Organization: UCAR/Unidata
>Keywords: 200303061342.h26Dg5B2003828 McIDAS ADDE XCD
Christian,
OK, I have made several changes on io that are related to LDM and McIDAS.
Here goes:
LDM (as 'ldm')::
1) backed up ~ldm/etc/ldmd.conf into ~ldm/etc/ldmd.conf.20030416
2) modified ~ldm/etc/ldmd.conf to group the several requests for products
together to make things easier to read
3) changed IDS|DDS|PPS to IDS|DDPLUS in both ~ldm/etc/ldmd.conf and
~ldm/etc/pqact.conf (for the McIDAS-XCD xcd_run line)
4) changed MCIDAS to UNIWISC in ~ldm/etc/ldmd.conf. This is the primary
name for the Unidata-Wisconsin feed in LDM-6. I did not, however,
change MCIDAS to UNIWISC in ~ldm/etc/pqact.conf. You may want
to do this some time.
5) added decoding entries for ANTARCTIC and global MOLLWEIDE imagery
in the Unidata-Wisconsin datastream. Since GEMPAK now supports
display of these images, I thought you might like to take advantage
of them in GEMPAK. Also added the '-a' and '-b' flags for each
pnga2area invocation in ~ldm/etc/pqact.conf (you had them for
some but not all)
6) verified that the ~ldm/etc/ldmd.conf entry for running McIDAS-XCD
decoders was OK. I then checked ~ldm/decoders/xcd_run to make
sure that all McIDAS-related environment variables were correct;
they were
7) changed ldm-mcidas decoding of NLDN lightning data from ~ldm/data/lightning
to ~ldm/data/mcidas. The reason for this was that McIDAS was setup
to look for the lightning data in the ~ldm/data/mcidas directory.
This affected the ability of McIDAS to use the data AND to scour
the data.
8) removed the scouring of ~ldm/data/mcidas by the LDM scour utility.
McIDAS files should be scoured by the McIDAS scouring script, mcscour.sh.
Since you have this setup in a cron entry for 'ldm', I figured I
could stop the scouring by 'scour'.
9) stopped the LDM
10) added a cron entry to rotate your ldm-mcidas log files
LDM-MCIDAS (as 'ldm')::
1) updated your ldm-mcidas decoders to v2002b:
cd ~ldm
ftp ftp.unidata.ucar.edu
<user> anonymous
<pass> your_full_email_address
cd pub/binary/irix_6.5-mips
get ldm-mcidas-2002b.tar.Z
quit
zcat ldm-mcidas-2002b.tar.Z | tar xvf -
rm ldm-mcidas-2002b.tar.Z
rm ldm-mcidas
ln -s ldm-mcidas-2002b ldm-mcidas
McIDAS (as 'mcidas')::
1) verified that the environment variable settings were OK
2) verified that the XCDDATA directory location was defined in the
McIDAS string table:
cd ~mcidas/workdata
tl.k XCDDATA
3) deleted the GROUPS.DAT, COUNTRY.DAT, CIRCUIT.DAT, and SIGCO.DAT files
in the ~mcidas/workdata directory. I did this so they would be
recreated from scratch in the next step.
I also deleted ~ldm/data/xcd/*.RA* files. I did this so that these
would be recreated in the next step.
4) ran the XCD setup BATCH files:
cd ~mcidas/workdata
batch.k XCD.BAT
batch.k XCDDEC.BAT
5) edited the ADDE dataset setup file, ~mcidas/data/LSSERVE.BAT. Since
I saw that you had a copy of this file in ~mcidas/workdata, I copied
it there as well. The changes I made were to define the elements
of the RTIMAGES and CIMSS datasets. I did this since you are
receiving these images and using them in GEMPAK. McIDAS can use
the files stored in the manner needed by GEMPAK. After updating
LSSERVE.BAT, I changed the DATALOC for CIMSS, RTIMAGES, RTPTSRC,
and RTWXTEXT to LOCAL-DATA so that I could define these datasets
on your machine (if you are pointing at a remote machine, the
DSSERVE commands in the LSSERVE.BAT file will give warnings and
not work). After doing this, I redefined ADDE datasets on
your machine:
cd ~mcidas/workdata
rm RESOLV.SRV
batch.k LSSERVE.BAT
6) edited the ADDE dataset pointer file, ~mcidas/data/LOCDATA.BAT.
I did this to point at your own system for the data that is being
decoded on it (e.g., RTIMAGES, CIMSS, RTPTSRC, RTWXTEXT), and pointed
at pscwx.plymouth.edu for NOAAPORT GINI images (as per a previous
email). After the editing, I made the DATALOCs active:
cd ~mcidas/workdata
batch.k LOCDATA.BAT
After steps 5) and 6), you are able to look at your own POINT data (e.g.,
surface, upper air, synoptic, ship/buoy, lightning) in McIDAS. You are
also able to look at imagery from the Unidata-Wisconsin datastream,
both GOES sectors and CIMSS products. In addition, since your ADDE
server is also setup, I can look at this same data on your system.
So, if you are running McIDAS on other systems at UQAM, or from user
accounts other than 'mcidas' on io, you may have to change your DATALOCs
to point at your own machine. Here are the McIDAS commands you would
use to point at the datasets you are now creating:
<in a McIDAS session>
DATALOC ADD CIMSS io.sca.uqam.ca
DATALOC ADD RTIMAGES io.sca.uqam.ca
DATALOC ADD RTPTSRC io.sca.uqam.ca
DATALOC ADD RTWXTEXT io.sca.uqam.ca
You can change those accounts use of datasets like GINICOMP, GINIEAST,
and GINIWEST to Plymouth State (for example) as follows:
DATALOC ADD GINICOMP pscwx.plymouth.edu
DATALOC ADD GINIEAST pscwx.plymouth.edu
DATALOC ADD GINIWEST pscwx.plymouth.edu
Also, you can access the NEXRAD Level III data and composites made from
them and McIDAS decoded NCEP model data from atm.geo.nsf.gov:
DATALOC ADD RTNEXRAD atm.geo.nsf.gov
DATALOC ADD NEXRCOMP atm.geo.nsf.gov
DATALOC ADD RTGRIDS atm.geo.nsf.gov
This is how I setup ~mcidas/data/LOCDATA.BAT, so you can look at it for
reference.
I think that things are all setup on io including scouring of the
McIDAS data files. Please let me know if you want more information on
anything I did, or see any problems with McIDAS XCD decoding or ADDE
serving.
Tom